Abstract

Modern memory systems are composed of several levels of caching. Design of these levels is largely an empirical practice. One highly-effective empirical method is the single-pass method wherein all caches in a broad design space are evaluated in one pass over the trace. Multiprogramming degrades memory system performance since (process) context switching reduces the effectiveness of cache memories. Few single-pass methods exist which account for multiprogramming effects. This paper uses a general model of single-pass algorithms, the recurrence/conflict model, and extends the model for recording the effects due to both voluntary context switches (e.g., system calls) and involuntary context switches (e.g., time quantum expiration). Involuntary context switches are modeled using the distribution of lengths between a reference to an address and the re-reference to the same address. The paper makes the assumptions that involunary context switches are equally likely to occur between each reference, and that one can independently estimate, fcs, the fraction of a cache’s contents flushed between context switches. The case where fcs= 1 is used to measure the effect of worst-case context switch penalty (the susceptibility) of several members of the SPEC89 benchmark set to context switching. Some empirical results of fcs are presented to illustrate the case where fcs< 1. The model is validated against its assumptions by comparing its results with more-restrictive methods.

Original languageEnglish (US)
Pages (from-to)994-1003
Number of pages10
JournalIEEE Transactions on Computers
Volume43
Issue number9
DOIs
StatePublished - Sep 1994

Fingerprint

Susceptibility
Switches
Switch
Multiprogramming
Computer systems
Time switches
Data storage equipment
Cache memory
Cache
Equally likely
Context
Caching
Model
Recurrence
Penalty
System Performance
Trace
Benchmark
Estimate

Keywords

  • Multiprogramming
  • SPEC
  • bench marking
  • cache
  • memory hierarchy
  • performance anaysis
  • simulation
  • single-pass algorithm

ASJC Scopus subject areas

  • Software
  • Theoretical Computer Science
  • Hardware and Architecture
  • Computational Theory and Mathematics

Cite this

The Susceptibility of Programs to Context Switching. / Hwu, Wen mei W.

In: IEEE Transactions on Computers, Vol. 43, No. 9, 09.1994, p. 994-1003.

Research output: Contribution to journalArticle

@article{dc18b7b89af04b489c54b6501fc89614,
title = "The Susceptibility of Programs to Context Switching",
abstract = "Modern memory systems are composed of several levels of caching. Design of these levels is largely an empirical practice. One highly-effective empirical method is the single-pass method wherein all caches in a broad design space are evaluated in one pass over the trace. Multiprogramming degrades memory system performance since (process) context switching reduces the effectiveness of cache memories. Few single-pass methods exist which account for multiprogramming effects. This paper uses a general model of single-pass algorithms, the recurrence/conflict model, and extends the model for recording the effects due to both voluntary context switches (e.g., system calls) and involuntary context switches (e.g., time quantum expiration). Involuntary context switches are modeled using the distribution of lengths between a reference to an address and the re-reference to the same address. The paper makes the assumptions that involunary context switches are equally likely to occur between each reference, and that one can independently estimate, fcs, the fraction of a cache’s contents flushed between context switches. The case where fcs= 1 is used to measure the effect of worst-case context switch penalty (the susceptibility) of several members of the SPEC89 benchmark set to context switching. Some empirical results of fcs are presented to illustrate the case where fcs< 1. The model is validated against its assumptions by comparing its results with more-restrictive methods.",
keywords = "Multiprogramming, SPEC, bench marking, cache, memory hierarchy, performance anaysis, simulation, single-pass algorithm",
author = "Hwu, {Wen mei W.}",
year = "1994",
month = "9",
doi = "10.1109/12.312110",
language = "English (US)",
volume = "43",
pages = "994--1003",
journal = "IEEE Transactions on Computers",
issn = "0018-9340",
publisher = "IEEE Computer Society",
number = "9",

}

TY - JOUR

T1 - The Susceptibility of Programs to Context Switching

AU - Hwu, Wen mei W.

PY - 1994/9

Y1 - 1994/9

N2 - Modern memory systems are composed of several levels of caching. Design of these levels is largely an empirical practice. One highly-effective empirical method is the single-pass method wherein all caches in a broad design space are evaluated in one pass over the trace. Multiprogramming degrades memory system performance since (process) context switching reduces the effectiveness of cache memories. Few single-pass methods exist which account for multiprogramming effects. This paper uses a general model of single-pass algorithms, the recurrence/conflict model, and extends the model for recording the effects due to both voluntary context switches (e.g., system calls) and involuntary context switches (e.g., time quantum expiration). Involuntary context switches are modeled using the distribution of lengths between a reference to an address and the re-reference to the same address. The paper makes the assumptions that involunary context switches are equally likely to occur between each reference, and that one can independently estimate, fcs, the fraction of a cache’s contents flushed between context switches. The case where fcs= 1 is used to measure the effect of worst-case context switch penalty (the susceptibility) of several members of the SPEC89 benchmark set to context switching. Some empirical results of fcs are presented to illustrate the case where fcs< 1. The model is validated against its assumptions by comparing its results with more-restrictive methods.

AB - Modern memory systems are composed of several levels of caching. Design of these levels is largely an empirical practice. One highly-effective empirical method is the single-pass method wherein all caches in a broad design space are evaluated in one pass over the trace. Multiprogramming degrades memory system performance since (process) context switching reduces the effectiveness of cache memories. Few single-pass methods exist which account for multiprogramming effects. This paper uses a general model of single-pass algorithms, the recurrence/conflict model, and extends the model for recording the effects due to both voluntary context switches (e.g., system calls) and involuntary context switches (e.g., time quantum expiration). Involuntary context switches are modeled using the distribution of lengths between a reference to an address and the re-reference to the same address. The paper makes the assumptions that involunary context switches are equally likely to occur between each reference, and that one can independently estimate, fcs, the fraction of a cache’s contents flushed between context switches. The case where fcs= 1 is used to measure the effect of worst-case context switch penalty (the susceptibility) of several members of the SPEC89 benchmark set to context switching. Some empirical results of fcs are presented to illustrate the case where fcs< 1. The model is validated against its assumptions by comparing its results with more-restrictive methods.

KW - Multiprogramming

KW - SPEC

KW - bench marking

KW - cache

KW - memory hierarchy

KW - performance anaysis

KW - simulation

KW - single-pass algorithm

UR - http://www.scopus.com/inward/record.url?scp=34547618891&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34547618891&partnerID=8YFLogxK

U2 - 10.1109/12.312110

DO - 10.1109/12.312110

M3 - Article

AN - SCOPUS:34547618891

VL - 43

SP - 994

EP - 1003

JO - IEEE Transactions on Computers

JF - IEEE Transactions on Computers

SN - 0018-9340

IS - 9

ER -